Common complaints of the elderly are poor sleep at night and excessive daytime sleepiness. Polygraphic studies of sleep reveal that the elderly have more Stage 1 and 2 NREMS, less Stage 3 and 4, lower delta power, and greater sleep fragmentation (more and longer arousals) than are seen in younger subjects. Daytime sleepiness in the elderly is more highly correlated with degree of sleep fragmentation than it is with total sleep time. This project will test hypotheses about causes of sleep structure changes with aging that result in sleep fragmentation, poor sleep quality, and deficits in daytime alertness. As in humans, the major changes in sleep structure in aged rats (our animal model) are increased sleep fragmentation and decreased consolidation of sleep in the inactive phase of the daily cycle. We employ a new technique and a new theory of sleep homeostasis in our research. The technique is a computer algorithm that uses the fast Hartley transform to do Fourier analysis of the EEG and a detect a complex of events always associated with the transition from NREMS to REMS (NRT) even if no scorable REMS occurs. This technique enables quantification of two critical variables of sleep structure -- REMS timing and maintenance of REMS. Our main hypothesis is that sleep fragmentation with age is due to deficits in mechanisms of REMS maintenance. Our theory of sleep homeostasis differs from convention by positing that REMS need builds up during NREMS, not during waking. Thus decreased REMS maintenance leads to increased REMS drive that disrupts NREMS by increasing the frequency of NRTs. The proposed research tests hypotheses about mechanisms contributing to poor REMS maintenance and sleep fragmentation in the aged rat. They include decline in cholinergic activity, decreased REMS drive secondary to decreased depth of NREMS, and deficits in the thermoregulatory system.